Volume 375, Issue 9709, Pages 123-131 (January 2010) Routine versus clinically driven laboratory monitoring of HIV antiretroviral therapy in Africa (DART): a randomised non-inferiority trial    The Lancet  Volume 375, Issue 9709, Pages 123-131 (January 2010) DOI: 10.1016/S0140-6736(09)62067-5 Copyright © 2010 Elsevier Ltd Terms and Conditions

Figure 1 Trial profile CDM=clinically driven monitoring. LCM=laboratory and clinical monitoring. ART=antiretroviral therapy. *Main reason for ineligibility. † 221 (55%) participants subsequently returned and were rescreened. ‡One participant had taken ART before starting the trial, and was not prescribed trial drugs or followed up; one individual without a consent form at monitoring defaulted before 8 weeks. The Lancet 2010 375, 123-131DOI: (10.1016/S0140-6736(09)62067-5) Copyright © 2010 Elsevier Ltd Terms and Conditions

Figure 2 Substitution in first-line ART and switch to second-line ART Percentages of participants still on first-line antiretroviral therapy (ART) and ever substituted to an alternate first-line regimen were estimated with cumulative incidences (deaths on first-line therapy were treated as competing risks, as was switch to second-line ART for first-line substitutions). Cause-specific hazard models in which deaths (and switch to second-line ART for first-line substitutions) are censored were used to calculate hazard ratios, stratified by randomisation factors. HR=hazard ratio. CDM=clinically driven monitoring. LCM=laboratory and clinical monitoring. The Lancet 2010 375, 123-131DOI: (10.1016/S0140-6736(09)62067-5) Copyright © 2010 Elsevier Ltd Terms and Conditions

Figure 3 Clinical disease progression (A) and adverse events (B) All hazard ratios were stratified according to randomisation factors, and p values were calculated with the log-rank test. Number needed to monitor for 1 year to avoid one (first) event was 130 (death) and 59 (new WHO stage 4 event of death) participants. Survival p values were 0·95 at 1 year, 0·92 at 3 years, and 0·90 at 5 years for laboratory and clinical monitoring (LCM) group; 0·94, 0·90, and 0·87 for clinically driven monitoring (CDM) group; and 0·55, 0·18, and 0·08 for the Entebbe15 cohort, respectively. HR=hazard ratio. ART=antiretroviral therapy. *Data from HIV-infected population of similar disease stage between 1996 and 2000. The Lancet 2010 375, 123-131DOI: (10.1016/S0140-6736(09)62067-5) Copyright © 2010 Elsevier Ltd Terms and Conditions

Figure 4 CD4-cell counts (A) Mean absolute CD4-cell count with time (unadjusted). CD4-cell counts were done every 12 weeks—small decreases at weeks 60, 84, and 108 are a result of the structured treatment interruption randomisation (terminated). Increase in mean CD4 count in the first 12 weeks was 102 cells per μL (95% CI 98–106) in clinically driven monitoring (CDM) group versus 103 cells per μL (99–107) in laboratory and clinical monitoring (LCM) group (p=0·77). Mean increase per year was subsequently 35 cells per μL (33–37) in CDM group and 42 cells per μL (40–44) in LCM group (p<0·0001). (B) Last CD4-cell count on first-line antiretroviral therapy (ART) or at switch to second-line ART. Number of deaths was 82 during the first year on first-line ART and 37 in the second year in LCM group and 97 during the first year on first-line ART and 32 in the second year in CDM group. The Lancet 2010 375, 123-131DOI: (10.1016/S0140-6736(09)62067-5) Copyright © 2010 Elsevier Ltd Terms and Conditions